Alzheimer disease (AD) is a leading cause of dementia in the elderly. Identifying susceptibility genes for AD will aid in risk assessment, diagnosis and understanding the etiology of the disease. Several chromosomal regions have been identified as harboring genes for late-onset AD, but only one gene, APOE, has been demonstrated consistently to be directly involved in disease risk. While APOE may explain up to half of the genetic effect in late-onset AD, the remaining susceptibility genes have been difficult to identify. A large number of candidate genes have been tested for association using either case-control tests in unrelated individuals or family-based association tests, but the results have varied greatly between studies. It is our hypothesis that epistatic effects among multiple genes play a more important role in determining risk of AD than the independent effects of any single gene. The multiple positive and negative reports for numerous candidate gene association studies may in fact be due to study designs that only evaluate each candidate gene for main effects that are independent of all other genes. The goal of this proposal is to determine the role of epistasis or nonadditive gene-gene interactions on the risk of late-onset AD. To accomplish this goal, we propose to identify and genotype several single nucleotide polymorphisms in ten AD candidate genes for which there have been conflicting association studies. We will use a family-based design using affected and unaffected siblings and propose to study 800 sibships containing approximately 1600 individuals. We will extend the recently developed pedigree disequilibrium test (PDT) to test for associations at the genotypic level and allow its incorporation into the new multifactor dimensionality reduction (MDR) method. To test for complex genetic interactions in these data, we will use this modified MDR-PDT method. We anticipate that these results will explain at least some of the inconsistency arising from Alzheimer disease candidate gene association studies. Further, knowledge gained from the proposed research will be invaluable for public health efforts to prevent and treat the initiation, progression, and severity of Alzheimer disease.